Clinical Trials are where International Banking was before SWIFT
When discussing clinical-trial documentation, most conversations revolve around repositories. How do we eliminate paper? How do we improve the eTMF? How do we connect sites, CROs, and sponsors? How do we automate filing and quality control?
These are reasonable questions. However, they implicitly assume that the primary challenge is document management. I am not convinced that this is the case.
The banking industry faced a remarkably similar situation during the late 1960s and early 1970s. International payments were growing rapidly, but the participating institutions were highly heterogeneous. Large international banks operated sophisticated data centres and large operations teams. Smaller regional banks often relied on local conventions, manual processing, telex, fax, and relatively small IT departments.
International payments already flowed between these institutions through correspondent banking relationships. A payment initiated by a local bank could pass through several intermediary banks before reaching the beneficiary bank. The ecosystem was therefore connected.
What it lacked was interoperability.
Each institution maintained its own systems, its own processes, and often its own interpretation of how international payments should be handled. Messages were reformatted multiple times, information was re-entered manually, and operational exceptions were common. As payment volumes increased, these inefficiencies became increasingly difficult to sustain.
The problem was not the absence of ledgers. Every bank already had a ledger. Nor was the problem a lack of ownership. Every bank remained responsible for its own customers, records, compliance obligations, and operations. The challenge was coordination between independent participants.
The response was SWIFT.
Looking back, it is tempting to view SWIFT primarily as a messaging network. In reality, it was something more fundamental. It created a common framework that allowed independent institutions to exchange information in a predictable manner while remaining fully autonomous. Banks continued competing on customers, products, service quality, and innovation, but gradually stopped competing on how payment information should be exchanged between institutions.
This distinction is important. SWIFT did not attempt to become a global bank. It did not centralise customer accounts. It did not replace the ledgers maintained by individual institutions. It acted as a trusted intermediary that facilitated communication between sovereign participants while allowing them to remain independent.
Fifty years later, clinical trials appear to exhibit many of the same characteristics.
A modern clinical trial involves sponsors, CROs, SMOs, investigative sites, laboratories, ethics committees, and technology providers. Each participant generates, reviews, stores, or depends upon documentation created by another participant. Each organisation operates its own systems, workflows, governance processes, and compliance procedures. And some participants may still rely on paper-based processes.
The industry has invested heavily in digitisation. Electronic Trial Master Files, Clinical Trial Management Systems, Electronic Data Capture platforms, site repositories, and collaboration portals are now commonplace. Compared to twenty years ago, the ecosystem is unquestionably more digital.
Yet digitisation and interoperability are not the same thing.
Sponsors still invest heavily in TMF review. CROs still perform extensive reconciliation activities. Sites continue to respond to documentation queries. Inspection-readiness programmes consume significant effort and budget. Despite substantial investment in digitisation, the industry’s ability to evaluate documentation remains heavily dependent on human interpretation.
This should make us pause.
If the primary challenge were simply digitisation, one might expect these activities to decrease over time. Instead, many organisations report increasing documentation volumes, increasing review effort, and increasing complexity. The industry has become more digital, but not necessarily easier to operate.
The TMF Reference Model has undoubtedly contributed to the industry’s maturity. It provides a common filing structure and a shared vocabulary for document organisation. This was an important step forward and has enabled a level of consistency that would have been difficult to imagine two decades ago.
However, the TMF Reference Model primarily addresses where documents belong. It does not establish a common interpretation of what those documents mean. A delegation log remains a delegation log regardless of where it is filed. An informed consent form remains an informed consent form regardless of which repository stores it. Filing location and document meaning are related, but they are not the same thing.
This distinction may be more important than it first appears.
Banking ultimately discovered that exchanging information electronically was only part of the challenge. The larger challenge was helping independent institutions arrive at a common understanding of the information they exchanged and process that information consistently. Connectivity turned out to be necessary, but not sufficient.
Clinical trials may be approaching a similar point.
The industry has made significant progress in digitising documentation. The next challenge may not be the creation of yet another repository or workflow platform. It may be deciding which aspects of documentation governance should remain competitive and which would benefit from a more common approach.
That is ultimately the question that interests me.
Not how documents are stored.
But how an ecosystem of independent participants arrives at a common understanding of what those documents demonstrate.
One Reply to “”